PROJECTSUMMARY?PROJECT1 ThepopulationofAmericanswithAlzheimer?sDiseaseandrelateddementia(ADRD)isexpectedtotripleby 2050.Thereisgreathopefornewtreatmentoptions,butcurrentlywearelimitedtomanagingestablished diseaseandavoidingfurthernegativeoutcomes.Whilenobreakthroughtreatmentshaveemerged,thelast20 yearshaveseenlargechangesinthediagnosisandmanagementofADRDwiththeintroductionofthefirst symptomaticdrugsandadvanceddiagnostictechnologies,suchasamyloidpositronemissiontomography (PET)scanningandcerebrospinalfluidbiomarkers.Whetherearlydiagnosisishelpfulisunknownand controversial,asevidencedbynationalguidelinesthatdisagreeonwhethertoscreenforADRD.Themixed messageslikelycontributetowidedifferencesinobservedstate-specificprevalenceofclinicallydiagnosed ADRDfromwhatwouldbeexpectedbasedonepidemiologicalestimates,rangingfrom30%belowto10% abovetheexpectedprevalence.Suchdifferencesinascertainmentofdiseasesuggestthattheprocessof gettingaclinicaldiagnosisofADRDvariesacrosshealthsystems.Usingthisvariationacrosshealthsystems andchangesovertime,weproposetogaininsightsintohowdiagnosticintensityandearlydiagnosisrelateto clinicaloutcomesofhospitalization,pharmaceuticaluse,andend-of-lifecareamongelderlypopulationsinthe U.S.WewillusenationalMedicareadministrativedatalinkedtonursinghomedata,theHealth&Retirement Studyfrom1996-2018andtheNurse?sHealthStudyfrom1995-2018.Ourunderlyinghypothesisisthatin someregions,peoplearebothmorelikelytoreceiveadiagnosisofADRDandmorelikelytobediagnosed early.Wewillusethesequasi-experimentalvariationsintheextentandtimingofdiagnosisacrossregionsand overtimetoassessthevalueofearlydiagnosisonsubsequenttreatmentsandclinicaloutcomes.